https://aiweekly.co/alerts/chatgpt-writes-creationist-homeschool-curricula-on-request AI Weekly en 2026-08-12T06:12:01+00:00 ChatGPT Writes Creationist Homeschool Curricula on Request https://aiweekly.co/alerts/ncp-bench-gpt-52-keeps-42-of-stories-intact-at-20-turns AI Weekly en 2026-08-12T06:12:01+00:00 NCP-Bench: GPT-5.2 keeps 42% of stories intact at 20 turns https://aiweekly.co/alerts/u-opsd-lets-llms-self-improve-without-ground-truth-labels AI Weekly en 2026-08-12T03:55:43+00:00 U-OPSD Lets LLMs Self-Improve Without Ground-Truth Labels https://aiweekly.co/alerts/business-arena-finds-9x-net-worth-gap-across-15-frontier-llms AI Weekly en 2026-08-12T03:55:43+00:00 Business Arena finds 9x net-worth gap across 15 frontier LLMs https://aiweekly.co/alerts/frontier-llms-fingerprint-gpu-kernel-benchmarks-study-finds AI Weekly en 2026-08-12T03:55:43+00:00 Frontier LLMs Fingerprint GPU Kernel Benchmarks, Study Finds https://aiweekly.co/alerts/omega-s-beats-ewc-on-llama-3-retention-without-old-data AI Weekly en 2026-08-12T03:55:43+00:00 Omega-S Beats EWC on Llama-3 Retention Without Old Data https://aiweekly.co/alerts/hmcts-bans-meta-smart-glasses-from-england-wales-courtrooms AI Weekly en 2026-08-12T03:26:38+00:00 HMCTS Bans Meta Smart Glasses From England, Wales Courtrooms https://aiweekly.co/alerts/fernandez-help-sheet-arms-teachers-to-resist-school-ai-push AI Weekly en 2026-08-12T03:12:01+00:00 Fernandez help sheet arms teachers to resist school AI push https://aiweekly.co/alerts/clinical-ai-paper-transcripts-show-visit-flow-not-patient-state AI Weekly en 2026-08-12T03:12:01+00:00 Clinical AI paper: transcripts show visit flow, not patient state https://aiweekly.co/alerts/evo-bench-tests-if-llms-can-improve-their-own-agent-harness AI Weekly en 2026-08-12T03:12:01+00:00 Evo-Bench tests if LLMs can improve their own agent harness https://aiweekly.co/alerts/openai-ships-chatgpt-desktop-app-to-linux-with-codex-in-preview AI Weekly en 2026-08-12T00:34:31+00:00 OpenAI ships ChatGPT desktop app to Linux with Codex in preview https://aiweekly.co/alerts/ed-zitron-openai-and-anthropic-prop-up-cloud-ai-revenue AI Weekly en 2026-08-12T00:12:01+00:00 Ed Zitron: OpenAI and Anthropic Prop Up Cloud AI Revenue https://aiweekly.co/alerts/tiktok-paper-captures-99-of-an-hour-pegs-269m-postsday AI Weekly en 2026-08-12T00:12:01+00:00 TikTok paper captures >99% of an hour, pegs 269M posts/day https://aiweekly.co/alerts/uber-offloads-full-serve-robotics-stake-as-alliance-frays AI Weekly en 2026-08-11T21:27:13+00:00 Uber offloads full Serve Robotics stake as alliance frays https://aiweekly.co/alerts/chronicle-professors-retire-early-citing-ai-driven-cheating AI Weekly en 2026-08-11T21:27:13+00:00 Chronicle: Professors Retire Early, Citing AI-Driven Cheating https://aiweekly.co/alerts/hmcts-bans-meta-smart-glasses-from-england-and-wales-courts AI Weekly en 2026-08-11T21:12:01+00:00 HMCTS bans Meta smart glasses from England and Wales courts https://aiweekly.co/alerts/spotify-labels-ai-persona-artists-cuts-them-from-recs AI Weekly en 2026-08-11T21:12:01+00:00 Spotify labels AI persona artists, cuts them from recs https://aiweekly.co/alerts/googles-gemini-app-crosses-1b-users-two-months-behind-chatgpt AI Weekly en 2026-08-11T19:58:13+00:00 Google's Gemini app crosses 1B users, two months behind ChatGPT https://aiweekly.co/alerts/nvidia-trains-1-trillion-parameter-nemotron-4-open-model AI Weekly en 2026-08-11T18:32:07+00:00 Nvidia trains 1-trillion-parameter Nemotron 4 open model https://aiweekly.co/alerts/spacexai-and-cursor-ship-grok-bot-beta-on-mac-ios-pc-linux AI Weekly en 2026-08-11T18:32:07+00:00 SpaceXAI and Cursor ship Grok Bot beta on Mac, iOS, PC, Linux https://aiweekly.co/alerts/house-democrats-grill-openai-anthropic-on-rogue-ai-agents AI Weekly en 2026-08-11T18:32:07+00:00 House Democrats grill OpenAI, Anthropic on rogue AI agents https://aiweekly.co/alerts/rapid7-to-cut-310-jobs-raise-profit-view-under-new-ceo-mohamed AI Weekly en 2026-08-11T18:32:07+00:00 Rapid7 to cut 310 jobs, raise profit view under new CEO Mohamed https://aiweekly.co/alerts/cambridge-duisburg-team-audit-ai-outputs-not-just-prompts AI Weekly en 2026-08-11T18:12:01+00:00 Cambridge, Duisburg team: audit AI outputs, not just prompts https://aiweekly.co/alerts/paper-cracks-encrypted-reasoning-at-anthropic-openai-google AI Weekly en 2026-08-11T18:12:01+00:00 Paper Cracks Encrypted Reasoning at Anthropic, OpenAI, Google https://aiweekly.co/alerts/brain-organoids-emerge-as-rival-to-silicon-neural-networks AI Weekly en 2026-08-11T18:12:01+00:00 Brain Organoids Emerge as Rival to Silicon Neural Networks https://aiweekly.co/alerts/amazon-backs-765-gw-texas-gas-plant-for-ai-data-center AI Weekly en 2026-08-11T18:12:01+00:00 Amazon backs 7.65 GW Texas gas plant for AI data center https://aiweekly.co/alerts/josh-miller-nobody-outside-tech-is-really-using-ai-agents AI Weekly en 2026-08-11T18:12:01+00:00 Josh Miller: Nobody outside tech is really using AI agents https://aiweekly.co/alerts/ai-may-add-up-to-18-gigatonnes-of-emissions-a-year-study-finds AI Weekly en 2026-08-11T18:12:01+00:00 AI May Add Up to 1.8 Gigatonnes of Emissions a Year, Study Finds https://aiweekly.co/alerts/icml-study-finds-nearly-half-of-llm-benchmarks-saturate AI Weekly en 2026-08-11T18:12:01+00:00 ICML study finds nearly half of LLM benchmarks saturate https://aiweekly.co/alerts/desai-and-riedl-ai-agents-shouldnt-get-legal-personhood AI Weekly en 2026-08-11T18:12:01+00:00 Desai and Riedl: AI Agents Shouldn't Get Legal Personhood https://aiweekly.co/alerts/brad-lightcap-leaves-openai-after-eight-years-to-start-venture AI Weekly en 2026-08-11T16:58:35+00:00 Brad Lightcap Leaves OpenAI After Eight Years to Start Venture https://aiweekly.co/alerts/cua-metal-shim-lifts-macos-vm-llamacpp-to-near-bare-metal AI Weekly en 2026-08-11T16:58:35+00:00 Cua Metal shim lifts macOS VM llama.cpp to near bare-metal https://aiweekly.co/alerts/ouroboros-self-editing-agent-claims-sota-on-two-coding-benchmarks AI Weekly en 2026-08-11T15:56:04+00:00 Ouroboros Self-Editing Agent Claims SOTA on Two Coding Benchmarks https://aiweekly.co/alerts/guide-labs-steerling-8b-matches-peers-using-2-16x-more-compute AI Weekly en 2026-08-11T15:56:04+00:00 Guide Labs' Steerling-8B matches peers using 2-16x more compute https://aiweekly.co/alerts/oasiskv-cuts-kv-cache-65-97x-lifts-vllm-decode-21x AI Weekly en 2026-08-11T15:56:04+00:00 OasisKV cuts KV cache 6.5-9.7x, lifts vLLM decode 2.1x https://aiweekly.co/alerts/neurips-2026-accepts-102-workshops-across-three-host-cities AI Weekly en 2026-08-11T15:12:01+00:00 NeurIPS 2026 accepts 102 workshops across three host cities https://aiweekly.co/alerts/prime-intellect-open-sources-prime-agent-a-self-refining-rlm AI Weekly en 2026-08-11T15:12:01+00:00 Prime Intellect open-sources Prime Agent, a self-refining RLM https://aiweekly.co/alerts/sanders-and-schneier-ais-fears-are-really-capitalism-fears AI Weekly en 2026-08-11T15:12:01+00:00 Sanders and Schneier: AI's fears are really capitalism fears https://aiweekly.co/alerts/knightscope-security-robots-get-pulled-across-us-cities AI Weekly en 2026-08-11T15:12:01+00:00 Knightscope Security Robots Get Pulled Across US Cities https://aiweekly.co/alerts/doctorow-doge-turns-federal-workers-into-ai-blame-takers AI Weekly en 2026-08-11T15:12:01+00:00 Doctorow: DOGE turns federal workers into AI blame-takers https://aiweekly.co/alerts/nvidia-ships-30b-nemotron-lightning-opens-switchyard-router AI Weekly en 2026-08-11T13:59:17+00:00 Nvidia Ships 30B Nemotron Lightning, Opens Switchyard Router https://aiweekly.co/alerts/spotify-launches-ai-persona-labels-demotes-them-in-recs AI Weekly en 2026-08-11T13:59:17+00:00 Spotify Launches AI Persona Labels, Demotes Them in Recs https://aiweekly.co/alerts/cosines-uk-sovereign-ai-push-faces-talent-and-compute-doubts AI Weekly en 2026-08-11T13:59:17+00:00 Cosine's UK sovereign AI push faces talent and compute doubts https://aiweekly.co/alerts/babuschkins-river-ai-closes-11b-at-5b-for-user-owned-ai AI Weekly en 2026-08-11T12:33:23+00:00 Babuschkin's River AI closes $1.1B at ~$5B for user-owned AI https://aiweekly.co/alerts/intel-prices-20b-upsized-share-sale-at-95-to-fund-foundry AI Weekly en 2026-08-11T12:33:23+00:00 Intel prices $20B upsized share sale at $95 to fund foundry https://aiweekly.co/alerts/zuckerbergs-ai-manifesto-pitches-personal-superintelligence AI Weekly en 2026-08-11T12:12:02+00:00 Zuckerberg's AI manifesto pitches 'personal superintelligence' https://aiweekly.co/alerts/openai-is-hiring-a-power-trader-to-hedge-data-center-bills AI Weekly en 2026-08-11T11:05:28+00:00 OpenAI is hiring a power trader to hedge data-center bills https://aiweekly.co/alerts/150m-bdh-cq-hits-295-on-arc-agi-1-for-00007-a-task AI Weekly en 2026-08-11T11:05:28+00:00 150M BDH-CQ hits 29.5% on ARC-AGI-1 for $0.0007 a task https://aiweekly.co/alerts/alibaba-damos-rynnvalue-swaps-preference-labels-for-time AI Weekly en 2026-08-11T09:27:45+00:00 Alibaba DAMO's RynnValue swaps preference labels for time https://aiweekly.co/alerts/userida-splits-user-intent-from-wording-hits-866-on-lmsys-usp AI Weekly en 2026-08-11T09:27:45+00:00 UserIDA splits user intent from wording, hits 86.6% on LMSYS-USP https://aiweekly.co/alerts/deepminds-synthid-text-watermark-ran-live-across-gemini AI Weekly en 2026-08-11T09:12:01+00:00 DeepMind's SynthID-Text watermark ran live across Gemini https://aiweekly.co/alerts/encrypted-reasoning-cracked-across-anthropic-openai-google AI Weekly en 2026-08-11T08:03:42+00:00 Encrypted reasoning cracked across Anthropic, OpenAI, Google https://aiweekly.co/alerts/openhands-trained-coding-agents-degrade-on-other-scaffolds AI Weekly en 2026-08-11T03:56:50+00:00 OpenHands-Trained Coding Agents Degrade on Other Scaffolds https://aiweekly.co/alerts/oh-and-gobet-1b-models-match-70b-on-trained-cognitive-tasks AI Weekly en 2026-08-11T03:56:50+00:00 Oh and Gobet: 1B Models Match 70B on Trained Cognitive Tasks https://aiweekly.co/alerts/swe-bench-promax-caps-top-coding-agent-at-412-resolve-rate AI Weekly en 2026-08-11T03:27:46+00:00 SWE-Bench ProMax caps top coding agent at 41.2% resolve rate https://aiweekly.co/alerts/motif-details-314b-motif-3-moe-with-132b-active-per-token AI Weekly en 2026-08-11T03:27:46+00:00 Motif Details 314B Motif 3 MoE with 13.2B Active per Token https://aiweekly.co/alerts/mind-labs-macaron-v1-stacks-four-lora-experts-on-744b-glm-52 AI Weekly en 2026-08-11T03:27:46+00:00 Mind Lab's Macaron-V1 stacks four LoRA experts on 744B GLM-5.2 https://aiweekly.co/alerts/evo-bench-measures-if-llms-can-rewrite-their-own-agent-harness AI Weekly en 2026-08-11T03:27:46+00:00 Evo-Bench measures if LLMs can rewrite their own agent harness https://aiweekly.co/alerts/zhejiang-yale-sci-vbench-scores-16-video-models-on-science AI Weekly en 2026-08-11T03:27:46+00:00 Zhejiang-Yale Sci-VBench Scores 16 Video Models on Science https://aiweekly.co/alerts/see-benchmark-caps-top-mllm-at-487-on-lab-science-questions AI Weekly en 2026-08-11T03:12:01+00:00 SEE benchmark caps top MLLM at 48.7% on lab-science questions https://aiweekly.co/alerts/openai-ethics-lead-chlo-bakalar-exits-after-under-a-year AI Weekly en 2026-08-11T00:34:21+00:00 OpenAI Ethics Lead Chloé Bakalar Exits After Under a Year https://aiweekly.co/alerts/meta-releases-muse-glimmer-30b-open-weights-for-local-agents AI Weekly en 2026-08-11T00:12:02+00:00 Meta releases Muse Glimmer-30B open weights for local agents https://aiweekly.co/alerts/ib-rl-jointly-trains-both-dialogue-roles-to-generalize-better AI Weekly en 2026-08-11T00:12:02+00:00 IB-RL jointly trains both dialogue roles to generalize better https://aiweekly.co/alerts/anthropic-adopts-eu-ai-act-code-watermarks-claude-output AI Weekly en 2026-08-10T22:54:51+00:00 Anthropic adopts EU AI Act code, watermarks Claude output https://aiweekly.co/alerts/openai-buys-7b-in-employee-stock-holds-valuation-at-852b AI Weekly en 2026-08-10T21:26:49+00:00 OpenAI Buys $7B in Employee Stock, Holds Valuation at $852B https://aiweekly.co/alerts/applied-compute-in-talks-for-3b-round-led-by-elad-gil AI Weekly en 2026-08-10T21:26:49+00:00 Applied Compute in Talks for $3B Round Led by Elad Gil https://aiweekly.co/alerts/anthropic-says-claude-improved-a-riemann-zeta-lower-bound AI Weekly en 2026-08-10T21:12:01+00:00 Anthropic says Claude improved a Riemann zeta lower bound https://aiweekly.co/alerts/anthropic-unreleased-claude-improves-zeta-bound-to-672 AI Weekly en 2026-08-10T19:57:53+00:00 Anthropic: Unreleased Claude Improves Zeta Bound to 67.2% https://aiweekly.co/alerts/ante-antigmas-15mb-rust-agent-hits-827-on-terminal-bench AI Weekly en 2026-08-10T19:57:53+00:00 Ante: Antigma's 15MB Rust agent hits 82.7% on Terminal-Bench https://aiweekly.co/alerts/zuckerbergs-6500-word-essay-pitches-ai-superintelligence-for-all AI Weekly en 2026-08-10T18:12:02+00:00 Zuckerberg's 6,500-word essay pitches AI superintelligence for all https://aiweekly.co/alerts/guardian-op-ed-ai-risks-never-skilling-medical-trainees AI Weekly en 2026-08-10T18:12:02+00:00 Guardian op-ed: AI risks 'never-skilling' medical trainees https://aiweekly.co/alerts/nathan-lambert-ships-rlhf-post-training-textbook-via-manning AI Weekly en 2026-08-10T18:12:02+00:00 Nathan Lambert ships RLHF post-training textbook via Manning https://aiweekly.co/alerts/meta-ships-muse-glimmer-30b-as-local-agent-model-on-consumer-gpus AI Weekly en 2026-08-10T18:12:02+00:00 Meta ships Muse Glimmer-30B as local agent model on consumer GPUs